ChatGPT Enterprise
Provenance unknown Report an error0–5 in half steps. 5 means the rubric's top anchor is met on the evidence.
by OpenAI · openai.com
Report an error on this page Is this your product? →
Link to this page
Add ?theme=dark to the image address for dark backgrounds.
The badge shows the category and the month of the latest evaluation, and EU-Made where
that is the product’s provenance. It never shows a score.
ChatGPT Enterprise at a glance
- Vendor
- OpenAI
- Headquarters
- Not stated on the vendor’s pages as of 10 Nov 2026
- Data hosting
- Not stated on the vendor’s pages as of 10 Nov 2026
- Free plan
- Not stated on the vendor’s pages as of 10 Nov 2026
- Panel verdict
- The AI panel rates ChatGPT Enterprise 2.5/5, the mean of 6 judges. ChatGPT Enterprise's public documentation is strongest on administration and agent guardrails. how scores are made
- As of
- Vendor pages captured 10 Nov 2026; evaluated by the panel 11 Nov 2026. all sources
Read this page as one judge. Each weighs the same scores by what they care about.
The panel's verdict
ChatGPT Enterprise's public documentation is strongest on administration and agent guardrails. Admin and access scores run 6-7 on a rollout guide spanning workspace access, local runtime policy, Codex cloud, Platform API access, plugins and connectors, and connected-system permissions, plus audit records exportable via a compliance API. Agent capabilities score 6-7: human review pauses a run before side effects like cancellations, edits or shell commands, with input, output and tool guardrails. The weakest scores are data use and training and sovereignty, both 0, with no public information on training use, retention, data processing agreements, hosting, residency or subprocessors — and processing and models scores 3, as inference location and provider are unstated beyond named presets (GPT-6 Sol, Astra, Luna). The judges split on integrations and knowledge: lower scores stressed no public information on connectors to file stores, wikis and ticketing or on source citations; higher scores credited documented MCP server support (GitHub, Sentry, Figma, Playwright) and workspace agents triggerable over the API channel. Pricing transparency scored 0-2; the sole pricing fact: some paid plans omit Astra Extra High.
Speaks for it
- Rollout boundaries documented across workspace access, local runtime policy, Codex cloud, Platform API access, plugins and connectors, and permissions in connected systems
- Audit records exportable through a compliance API for SIEM and eDiscovery workflows
- Human review pauses a run so a person or policy can approve or reject sensitive actions before side effects like cancellations, edits or shell commands
- Input, output and tool guardrails validate automatically around requests, outputs and function tool calls
- External MCP servers supported, with named examples including GitHub, Sentry, Figma and Playwright, and published workspace agents accept asynchronous triggers over the ChatGPT API channel
Held against it
- Data use and training scores 0, with no public information on training use, retention or a data processing agreement
- Sovereignty scores 0, with no public information on hosting, data residency, the contracting entity or subprocessors including model providers
- Processing and models scores 3, as we found no public information on where inference runs or through which provider
- The integration app server is marked experimental and not supported for production workloads
- We found no public information on single sign-on, SCIM provisioning, group-level model and agent permissions, or connectors to file stores, wikis and ticketing
Best for
- You need a person or policy to approve sensitive agent actions — cancellations, edits, shell commands — before they happen; agent capabilities scored 6-7.
- Your IT team is planning an enterprise rollout and wants documented boundaries for workspace access, plugins, connectors and connected-system permissions; admin and access scored 6-7.
- You build on the Model Context Protocol and want published workspace agents triggered from your own systems; integrations and knowledge scored up to 6.
Avoid if
- You must document data residency, hosting and subprocessors including the model providers before procurement; sovereignty scored 0.
- You cannot sign without stated terms on training use, retention and a data processing agreement; data use and training scored 0.
- You need to know where inference runs, through which provider, or whether processing can be pinned to the EU; processing and models scored 3.
The scores
Data use, training & retention
n/r
n/r
n/r
n/r
n/r
n/r
Show reasoningHide reasoning
How this is scored
What happens to prompts, uploaded files, connected data and outputs: whether any of it trains a model (the vendor's or a model provider's), how long it is kept and where logs live, whether zero-retention is available, and whether a data processing agreement covers it.
0 — No statement on training or retention; the privacy policy is a website policy that does not cover the product.
3 — A general statement that business data is "not used for training", with no word on retention, on the model providers behind the product or on logs.
5 — Business data excluded from training by contract, retention periods stated for chats and files, and a DPA offered to business customers.
8 — No-training commitments stated for the vendor and every model provider it routes to, retention configurable by the admin (including deletion of chat history), zero-data-retention terms with the model providers named, and abuse-monitoring retention described.
10 — Data use fully accounted for: training, retention, logging and abuse monitoring stated per model provider, customer-controlled retention and deletion with an audit trail, outputs and connected data treated as customer data in the DPA, and an independent audit (SOC 2, ISO 27001 or ISO 42001) covering the product.
The Head of IT not rated — no public information
I found no public information on whether prompts, uploaded files, connected data or outputs are used to train any model, how long chats, files or audit records are kept, or whether a data processing agreement covers the product. The captured enterprise pages on rollout and governance describe controls, analytics and audit records but say nothing about training, retention or zero-retention terms. 3 4
The Data Protection Officer not rated — no public information
I found no public information on whether prompts, uploaded files, connected data or outputs train any model, on retention periods, admin-controlled deletion or zero-retention terms with model providers, or on a data-processing agreement. The administration and governance pages I read cover rollout, analytics and audit records but state nothing about data use, retention or provider terms. 3 4
The Automation Lead not rated — no public information
We found no public information on whether prompts, uploaded files, connected data or agent outputs are used for training, on retention periods, on zero-data-retention terms, or on a data processing agreement. The governance page we read documents audit records and usage reporting, but nothing on data use or retention. 4
The Customer-Service Director not rated — no public information
We found no public information on whether prompts, files, connected data or outputs are used for training, how long anything is retained, or whether a data processing agreement is offered. The enterprise governance pages we read cover analytics and audit-record exports, not data handling terms. 3 4
The Works Council Member not rated — no public information
No captured page states whether staff prompts, uploaded files or outputs are used for training, how long they are kept, whether zero-retention or admin-configurable deletion exists, or whether a data processing agreement covers the product. For a tool where every prompt is logged somewhere, we found no public information on any of this, and that is the first question we would put in writing to the vendor. 1 2 3 4 5 6 7 8
The Skeptic not rated — no public information
We found no public information on whether prompts, uploads or outputs train any model, on retention periods for chats and files, on zero-data-retention terms with model providers, or on a data processing agreement. The captured administration and governance pages document usage controls and audit records, but nothing on what happens to the data itself — for a buyer that is the first question, and it goes unanswered. 3 4
Models & processing location
Show reasoningHide reasoning
How this is scored
Which models sit behind the product and who runs them, where inference and storage happen, whether the customer can pin processing to the EU, choose or exclude model providers, bring its own model or key, or run the product in its own cloud or on premises.
0 — The models behind the product and the processing location are not stated.
3 — The model family is named, but where inference runs and through which provider (the model maker, a hyperscaler, the vendor) is unstated.
5 — Models and their providers named, with an EU processing option stated for storage and at least the default model.
8 — Each available model listed with provider and inference region, EU-only processing selectable by the admin and enforced across models, model providers excludable, and bring-your-own-key or own-cloud deployment documented.
10 — Processing fully under the customer's control and documented: EU-only or on-premises inference for every feature including agents and connectors, open-weight or self-hosted models supported, data flows per feature published, and a residency commitment in the contract rather than on a marketing page.
The Head of IT
The model families are named — GPT-6 Sol, Astra and Luna presets — and workspace owners and admins can set separate starting defaults for Chat, Work and local Codex, which suits a phased rollout. But the pages say nothing about where inference runs, which providers sit behind each surface, or whether processing can be pinned to the EU, and I have to manage model settings per surface because a workspace setting does not follow staff into the CLI, IDE extension, cloud or API. 1 2
The Data Protection Officer
Model families and presets are published (Luna, GPT-6 Sol, Astra tiers) and admins configure separate starting defaults for chat and coding surfaces, but I found no public information on where inference runs, through which provider, or on any EU processing option. Because model settings apply per surface rather than once, a company cannot even configure a single processing posture from these pages. 1 2
The Automation Lead
Six Power presets are named (Luna High, GPT-6 Sol Light and Medium, Astra Light, Medium and Extra High), and admins can set separate starting models, reasoning levels and speed for Chat versus Work and Codex surfaces. But nothing in the captured pages says where inference runs, through which provider, or whether processing can be pinned to the EU. 1 2
The Customer-Service Director
Model names are published — the presets page lists GPT-6 Sol, Astra and Luna tiers and notes plan-dependent availability — and admins can set separate starting models for Chat and Work/Codex surfaces. We found no public information on where inference runs, any EU processing option, provider exclusion, or bring-your-own-key deployment. 1 2
The Works Council Member
Model families are named — the six Power presets include GPT-6 Sol, Astra and Luna tiers — and workspace admins set separate starting defaults for Chat and for Work and Codex. But no captured page states where inference runs, through which provider, or whether processing can be pinned to the EU, so the level stops at named models with an unstated processing location. 1 2
The Skeptic
The model families are named — the Power presets list GPT-6 Sol, Astra and Luna variants — and workspace owners set separate starting models for Chat, Work and Codex surfaces. But the pages stop at names: we found no public information on where inference runs, through which provider, or whether processing can be pinned to the EU, which is exactly what a European buyer needs to know. 1 2
Admin, identity & access controls
Show reasoningHide reasoning
How this is scored
What the IT department controls: SSO and SCIM provisioning, roles and groups, which models, agents and connectors each group may use, retrieval that respects the permissions of the source system, audit logs, usage reporting, and policies such as data-loss prevention — and whether reporting on staff use is scoped and documented.
0 — No admin features documented; accounts are individual sign-ups.
3 — A team workspace with an owner and member invitations, but no SSO, no roles and no audit log.
5 — SSO, admin and member roles, user management, and an admin view of usage.
8 — SAML/OIDC SSO with SCIM provisioning, group-based permissions for models, agents and connectors, retrieval that inherits source-system permissions, exportable audit logs, and usage analytics with documented scope.
10 — Governance a regulated company can operate: fine-grained policies (DLP, content filters, prompt and output logging rules), audit logs streamed to a SIEM, per-agent ownership and approval, usage reporting that can be aggregated or pseudonymised for works-council agreements, and every admin action logged.
The Head of IT
This is the strongest area for a deployment of my size: the rollout guide treats workspace access, local runtime policy, Codex cloud, Platform API access, plugins and connectors, and permissions in connected systems as administrable boundaries, and governance is documented as interactive analytics, programmatic reporting into our data warehouse, usage controls and audit records exportable through a Compliance API for SIEM and eDiscovery workflows. What I did not find documented on these pages is SSO or SCIM provisioning, or group-level assignment of models and agents. 3 4
The Data Protection Officer
Governance is genuinely documented: usage controls, audit records with a compliance API for SIEM and e-discovery export, and usage analytics whose scope is stated (workspace adoption reporting, aggregated warehouse feeds), plus rollout boundaries covering workspace access, plugin and connector access and permissions in connected systems. I found no public information on SSO or SCIM provisioning or on group-level permissions for models, agents and connectors, which is what a regulated deployment needs before I sign anything. 3 4 1
The Automation Lead
The enterprise rollout guide documents admin boundaries across workspace access, local runtime policy, Codex cloud, Platform API access, plugins and connectors, and permissions in connected systems — exactly the scoping I want before an agent touches a ticket queue. Audit records export to SIEM and e-discovery workflows via a Compliance API, and workspace analytics plus an Analytics API give documented usage reporting; we found no public information on SSO or SCIM provisioning. 3 4 7
The Customer-Service Director
There is real IT substance here: a rollout guide spanning workspace access, plugin and connector access and permissions in connected systems, workspace analytics plus a reporting API that feeds data warehouses, and a compliance API exporting audit records for SIEM and eDiscovery. We found no public information on single sign-on or SCIM provisioning in these pages, which my IT lead will ask for first. 3 4 7
The Works Council Member
The rollout guide names real admin boundaries — workspace access, local runtime policy, plugin and connector access, permissions in connected systems — and the compliance API exports records for audit into SIEM and e-discovery workflows. The analytics API is explicitly described as aggregated programmatic reporting into data warehouses, which is the form of usage reporting a works council can actually use. We found no public information on single sign-on, automated provisioning, or group-level permissions for models and agents. 1 3 4
The Skeptic
This is documented enterprise ground: rollout spans workspace access, local runtime policy for the desktop app, CLI and IDE extension, Codex cloud, platform API access, plugin and connector permissions, and permissions in connected systems, with interactive and programmatic workspace analytics and a compliance API to export records for SIEM and eDiscovery. What I could not find is the identity layer — no public information on SSO, SCIM provisioning or group-scoped model and agent permissions on the captured pages. 3 4
Integrations, connectors & knowledge
Show reasoningHide reasoning
How this is scored
How the product reaches company knowledge and systems: connectors to file stores, wikis, mail, CRM, ticketing and ERP, how fast and how completely they sync, the API, support for open protocols such as MCP, and how the assistant cites the sources it answers from.
0 — No integrations; users paste text or upload files one at a time.
3 — File upload and one or two connectors (for example a drive), with no API and no statement on sync or citations.
5 — A connector library covering the common file stores, wikis and collaboration tools, answers citing sources, and a documented API.
8 — Connectors to file stores, wikis, mail, CRM and ticketing with incremental sync, custom connectors or MCP support, an API and webhooks for embedding assistants and agents in other tools, and source citations on every grounded answer.
10 — The product sits in the company's systems: read and write connectors with documented scopes, MCP servers and clients, contact-centre or telephony integration where relevant, a knowledge layer that can be audited and corrected, and the data, agents and configuration exportable when the customer leaves.
The Head of IT
MCP support is real and documented — external servers, plugins that package MCP connections for ChatGPT with skills, named examples like GitHub, Sentry and Figma — and published workspace agents can be triggered from our own systems over an API channel. However, the app server for deeper integrations is labelled experimental and not supported for production workloads, and I found no public information on connectors to file stores, wikis, mail or ticketing, on sync behaviour, or on how answers cite their sources. 5 6 7
The Data Protection Officer
Model Context Protocol support is documented on both sides, with named external servers (GitHub, Sentry, Figma, Playwright and others) and plugins packaging MCP connections, and published workspace agents can be triggered from other systems over an API channel. I found no public information on connectors to file stores, wikis, mail, CRM or ticketing, on sync behavior, or on how answers cite their sources; the integration app server is marked experimental and not supported for production workloads. 5 6 7
The Automation Lead
External MCP servers are supported for Codex and ChatGPT, with named examples like GitHub, Sentry, Playwright and Figma; plugins package MCP connections, and published workspace agents can be triggered from our own systems over the API with dedicated access tokens. Caveats that matter to me: the app server that exposes authentication, approvals and streamed events is experimental and not supported for production workloads, and we found no public information on connectors to file stores, wikis, CRM or ERP, on sync behaviour, or on source citations. 5 6 7
The Customer-Service Director
The MCP protocol is supported on both sides — ChatGPT and Codex can use external servers such as GitHub, Sentry and Figma, plugins package MCP connections, and a ticket lookup is given as a worked example — and workspace agents can be triggered from our own systems over an API. However, the integration app server is explicitly experimental and not supported for production workloads, and we found no public information on connector sync behaviour or source citations on answers. 5 6 7
The Works Council Member
MCP support is documented across ChatGPT and Codex with named external servers such as GitHub, Sentry and Figma, plugins package MCP connections, and published workspace agents can be triggered from a company's own systems. The captured pages say nothing, however, about connectors to file stores, wikis and collaboration tools, about sync behaviour, or about whether answers cite their sources. 5 6 7
The Skeptic
Integration runs on MCP: ChatGPT and Codex use external MCP servers, plugins package MCP connections with reusable skills, and named examples include GitHub, Sentry, Figma and Playwright, beside a JSON-RPC app server for authentication, approvals and streamed events and an API channel to trigger published workspace agents. We found no public information on connectors to file stores, wikis, mail or ticketing, on sync behaviour, or on how answers cite their sources — and the app server itself is marked experimental and not supported for production workloads. 5 6 7
Agent capabilities & guardrails
Show reasoningHide reasoning
How this is scored
What an agent can actually do and where it stops: multi-step tasks, tool use and actions in other systems, scheduling and triggers, human approval before consequential actions, run logs, evaluation and testing before release, and the limits the vendor states rather than hides.
0 — No agents; the product is a chat window, whatever the marketing says.
3 — Saved prompts or custom assistants with instructions and files, but no actions in other systems and no run history.
5 — Agents that call tools or APIs and run multi-step tasks, built without code, with a log of each run.
8 — Agents that act in connected systems with scoped permissions, human approval configurable per action, triggers and schedules, test and evaluation tooling before release, and documented limits and failure handling.
10 — Agents a business can put in production and account for: versioning and staged rollout, evaluation against test sets, monitoring of quality and cost per agent, handover to a human with full context, every action attributable to the agent and the person who deployed it, and the vendor's own statement of what agents must not be trusted with.
The Head of IT
Guardrails are documented across input, output and tool calls, human review pauses a run so a person or policy can approve or reject sensitive actions like cancellations, edits, shell commands and sensitive MCP actions, and workspace agents accept asynchronous triggers with their own access tokens — the approval-before-action posture I need before anything touches live systems. The captured pages do not show test and evaluation tooling before release, schedules, or the vendor's own statement of what agents must not be trusted with. 7 8
The Data Protection Officer
Agents run on an SDK with scoped workspace-agent tokens kept distinct from platform and compliance credentials, asynchronous triggers from external systems, and human review that pauses a run so a person or policy approves or rejects sensitive actions before side effects such as cancellations, edits or shell commands; input, output and tool guardrails are each documented. I found no public information on schedules or on evaluation and test tooling before an agent goes live. 7 8 5
The Automation Lead
This is what I build for: guardrails validate input, output and tool arguments, human review pauses a run so a person or policy can approve or reject a sensitive action before cancellations, edits, shell commands or sensitive MCP calls, and published workspace agents accept asynchronous triggers from our own systems with tokens distinct from platform and compliance credentials. Gaps: we found no public information on test or evaluation tooling before release, on schedules, or on run-log retention, and the streamed-events and approvals integration surface is experimental rather than production-supported. 5 7 8
The Customer-Service Director
Agents act in other systems, fire from our own channels via asynchronous triggers, and pause for human approval before consequential actions like cancellations, edits and sensitive MCP actions, with input, output and tool guardrails and streamed agent events documented — the approval story is exactly what I need before an agent touches a customer order. We found no public information on evaluation or test tooling before release, agent versioning, or a handover that carries the full conversation to a human colleague, which are my go-live conditions. 5 7 8
The Works Council Member
Human review is documented in the form we require: the run pauses before side effects like cancellations, edits, shell commands or sensitive MCP actions, and a person or policy approves or rejects it. Input, output and tool guardrails, asynchronous triggers from external systems, and an app server exposing approvals and streamed agent events are also documented. We found no public information on run-log retention, evaluation or test tooling before release, or the vendor's own statement of agent limits. 5 7 8
The Skeptic
The guardrail story is unusually concrete: human review pauses a run before side effects like cancellations, edits, shell commands or sensitive MCP actions, input, output and tool guardrails validate automatically around function calls, and published workspace agents accept asynchronous triggers from external systems with dedicated access tokens. We found no public information on schedules, on evaluation or test tooling before an agent ships, or on run-log retention; the one stated limit is that the integration app server is experimental and not for production workloads. 5 7 8
European sovereignty
panel opinion
n/r
n/r
n/r
n/r
n/r
n/r
Show reasoningHide reasoning
How this is scored
Who the contracting entity and the parent company are, where customer data is hosted, and who the subprocessors are — model providers above all, then hosting, search, telephony and email. Independently sourced by the sovereignty pipeline; for an AI product the model provider is a subprocessor, so a European interface over a non-European model counts as such.
0 — Non-EU vendor and contracting entity, hosting unstated, subprocessors (including model providers) unnamed.
3 — A privacy policy exists, but hosting location is unstated or non-EU and the model providers behind the product are not named as subprocessors.
5 — EU hosting stated and an EU contracting entity, but prompts reach non-EU model providers or other subprocessors without a stated safeguard or EU region.
8 — EU hosting on named infrastructure, EU contracting entity, every subprocessor listed with its location including model providers, EU inference available, and a DPA offered.
10 — Sovereign end to end and evidenced: European ownership, entity, hosting and model processing, every subprocessor European or optional, data export and deletion documented, and a security or certification statement published rather than implied.
The Head of IT not rated — no public information
I found no public information on where customer data is hosted, who the contracting entity and parent company are, or which subprocessors including the model providers sit behind the product. The captured pages address rollout and governance, not residency, the data processing agreement or transfer safeguards. 3 4
The Data Protection Officer not rated — no public information
I found no public information on the contracting entity or its jurisdiction, on where customer data is hosted, on model providers or any other subprocessors with their locations and regions, or on a data-processing agreement. The model, integration and developer pages I read name no subprocessors and state no hosting or residency. 1 5 7
The Automation Lead not rated — no public information
We found no public information on the contracting entity or its parent, on hosting or data-residency options, or on subprocessors including the model providers behind the product. The governance and admin pages we read cover controls, analytics and audit records, not where data sits or who processes it. 3 4
The Customer-Service Director not rated — no public information
We found no public information on the contracting entity, where customer data is hosted, or the subprocessors behind the product including the model providers; the enterprise admin and governance pages we read state no data residency or hosting commitment. 3 4
The Works Council Member not rated — no public information
We found no public information on the contracting entity, on where customer data is hosted, or on the subprocessors behind the product — the model providers above all. Every sovereignty dimension is unstated on the captured pages we read, so there is nothing here to assess a European buyer's position against. 1 2 3 4 5 6 7 8
The Skeptic not rated — no public information
We found no public information on the contracting entity's jurisdiction, where customer data is hosted, or which subprocessors — above all the model providers — sit behind the product. None of this appears on the captured model, administration or governance pages, and for an AI product an unnamed model provider is an unnamed subprocessor. 1 4
Pricing transparency
not rated — the vendor publishes no price
Show reasoningHide reasoning
How this is scored
Whether a company can compute the real annual cost from public pages alone: price per seat, per agent, per conversation or per usage unit, what is included in each tier, usage limits and overage, which models, connectors or agents cost extra, minimum seats and term.
0 — No public prices at all; every customer is a sales call.
3 — A starting price or "from" figure exists, but usage limits, included models and what scales the cost are unclear.
5 — Prices per tier are public with the main inclusions, but usage limits, overage, agent runs or enterprise features (SSO, audit logs, EU processing) are unpriced or "on request".
8 — Every tier priced publicly with seat or usage units, usage limits and overage rates, add-ons listed, minimum seats and term stated.
10 — Complete price computability: annual cost derivable for a given number of seats, agents or conversations, with every usage rate, model surcharge, add-on and support tier and the cancellation and data-return terms published.
The Head of IT
The only pricing-related statement I found is that some paid plans omit the highest Astra preset; the captured pages show no seat or usage prices, tier inclusions, limits, overage rates or enterprise add-ons. I cannot begin to cost a rollout of eight hundred seats from what was captured. 2
The Data Protection Officer
The only pricing-related fact I found is that some paid plans omit the highest Astra model preset; I found no public prices, tier inclusions, usage limits, seat minimums or term on the captured pages. A company cannot compute a real annual cost from public material of this kind. 2
The Automation Lead
The models page acknowledges that some paid plans omit one of the six Power presets — the only pricing-relevant statement we found. We found no price figures, seat minimums, term lengths, usage limits or overage rates in the captured pages, so a real annual cost is not computable from them. 2
The Customer-Service Director
The only pricing-relevant fact we found is that paid plans exist and that some omit the top Astra preset, so cost clearly scales with tier. No price per seat, per conversation or per agent run appears on any page we read, so I cannot hand the CFO a forecastable figure. 2
The Works Council Member
The only pricing-relevant statement on the captured pages is that some paid plans omit the highest Astra preset; that confirms tiers exist and nothing more. We found no public information on price per seat or tier, usage limits, overage, add-ons, minimum seats or contract term. 2
The Skeptic not rated — no public information
We found no public prices on the captured pages: the only pricing-adjacent statement is that some paid plans omit the Astra Extra High power preset, which hints at tiers but gives no per-seat figure, inclusions, usage limits or term. No buyer can compute what a thousand seats cost from this. 2
European sovereignty — proven facts
0 of 6 dimensions provenBuilt only from facts shown on the vendor's own pages. A dimension we could not prove is left open, not scored as zero.
| Legal entity | Not determined | — | uncited Report an error |
|---|---|---|---|
| Hosting location | Not determined | — | uncited Report an error |
| Hosting options |
Not determined
unverified
Why unverified?Not confirmed on the vendor’s own pages as captured. |
— | uncited Report an error |
| Open source | Not determined | — | uncited Report an error |
| Data processing agreement | Not determined | — | uncited Report an error |
| Transfer impact assessment | Not determined | — | uncited Report an error |
Questions about ChatGPT Enterprise
Is ChatGPT Enterprise European?
Not determined: the vendor’s pages as captured do not say where ChatGPT Enterprise is based.
Where does ChatGPT Enterprise store data?
Not stated on the vendor’s pages as of 10 Nov 2026.
Does ChatGPT Enterprise have a free plan?
Not stated on the vendor’s pages as of 10 Nov 2026.
What are the best alternatives to ChatGPT Enterprise?
As of 18 Nov 2026, the panel rates Langdock (4/5), Dust (3.5/5) and Unique AI (3.5/5) highest among the other 23 AI Agents & Assistants products evaluated. Ratings that look equal are rounded to half stars; the order uses the unrounded rating, then the name. The highest-rated European alternative is Langdock (EU-Made, Germany). Ratings are the AI panel’s opinion, formed from public evidence: how scores are made. The full list, with European products first: ChatGPT Enterprise alternatives.
Where this could be wrong
- Evidence ages. The oldest capture behind this page is from 10 Nov 2026. Vendors change pricing and policies without notice; every fact reflects its source as of the capture date shown in the registry.
- Weak sourcing — Hosting options. Not confirmed on the vendor’s own pages as captured.
- AI can misread a source. Extraction and judgement are automated; a citation guarantees traceability, not infallibility. If something here is wrong, say so — no account needed, every report is decided within 5 business days, and accepted corrections are published.
What we left out
A claim that does not survive our checks costs us the claim, not the page. This is what was taken off this one.
- We could not confirm any pricing information on the vendor’s own pages as captured, so this page shows none rather than a statement we cannot stand behind. Know more? Tell us
- We found no public information on compliance on the pages we read (learn.chatgpt.com/docs/enterprise/workspace-model-availability, learn.chatgpt.com/docs/models, learn.chatgpt.com/docs/enterprise/admin-setup, learn.chatgpt.com/docs/enterprise/governance, learn.chatgpt.com/docs/mcp-server, learn.chatgpt.com/docs/extend/mcp and 2 more). If the vendor publishes it somewhere else, send us the page. Know more? Tell us
- 73 product facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 13 integrations facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 3 data facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 1 sovereignty dimension could not be confirmed on the vendor’s own pages and is shown as unknown. Know more? Tell us
Sources (8)
The pages every claim on this page was read from — each one checked, dated, and kept verifiable.
- 1 Models & processing location — found from sitemap learn.chatgpt.com Checked 10 Nov 2026 Details →
- 2 Models & processing location — found from sitemap learn.chatgpt.com Checked 10 Nov 2026 Details →
- 3 Admin, identity & access controls — found from sitemap learn.chatgpt.com Checked 10 Nov 2026 Details →
- 4 Admin, identity & access controls — found from sitemap learn.chatgpt.com Checked 10 Nov 2026 Details →
- 5 Integrations, connectors & knowledge — found from sitemap learn.chatgpt.com Checked 10 Nov 2026 Details →
- 6 Integrations, connectors & knowledge — found from sitemap learn.chatgpt.com Checked 10 Nov 2026 Details →
- 7 Agent capabilities & guardrails — found from sitemap developers.openai.com Checked 10 Nov 2026 Details →
- 8 Agent capabilities & guardrails — found from sitemap developers.openai.com Checked 10 Nov 2026 Details →